Inspiration

I wanted to try out a machine-learning related project this year, and this was one of the more beginner-friendly project which was perfect ! Using historical data and machine learning algorithms, we are able to predict the likelihood of an outbreak.

Influenza Outbreak Prediction

Developed for Code Without Barriers's hackathon, this project aims to predict influenza outbreaks in different regions based on historical data and various factors by analyzing previous patterns and trends in the provided dataset.

Dataset

The dataset used, "clinical-sentinel-laboratory-influenza-and-other-respiratory-virus-surveillance-data-by-region-and-influenza-season.csv," provides past information on the percentage of cases tested positive and is affected by variables such as season, date, region, etc.

Since there was no target variable, I used the following code to determine if the number of cases is considered an outbreak:

# define threshold to determine yes/no outbreak, in this case i set if >= 100 then considered outbreak
df["Outbreak"] = df["Number_Positive"].apply(lambda x: 1 if x >= 100 else 0)

How to use

To clone and run this application, you will need Git:

1. Clone the repository:

$ git clone https://github.com/mlemxy/Influenza-Prediction

# create virtual env
py -m venv .venv
.venv\scripts\activate

# install dependencies
$ pip install -r requirements.txt

Built With

• machine-learning
• python